Hacker News new | past | comments | ask | show | jobs | submit login
GM recalling nearly 69,000 Bolt EVs for fire risks (reuters.com)
176 points by hhs on Nov 13, 2020 | hide | past | favorite | 104 comments



This reminds me of the "Mahk from Haverhill" videos that parodied Chevy's 2017 TV commercials [0]

I really, really hope they crank out a few parodies for this recall.

[0] https://www.youtube.com/watch?v=d03svtYlFm8&list=PLoC1ZyvZc4...


I had never seen this before. Freaking hilarious!


Musk could probably make billions just by licensing their battery technology. Once EVs are truly mass market, competition, as with current automobiles, will drive down margins significantly. Licensing his tech at the same time would really boost the overall profit.

Then again I'm just arm-chairing this thing. It's not like I have my own billions to backup my business prowess.


There were multiple Tesla fires, and they nerfed barriers as a response, without issuing recall (https://electrek.co/2019/10/04/tesla-software-updat-battery-...)


I’m now old. Nerfed: The term "nerfing" comes from the online gaming world of Ultima Online.

At one point in the game, the developers reduced the power of swords in melee combat.

This resulted in players complaining that it was like they were hitting each other with nerf bats, not swords.

From then on, if ever something gets made less worth while than it had been originally, it is considered 'nerfed'

From https://www.urbandictionary.com/define.php?term=nerfed


Pretty sure Nerf preceded Ultima Online https://en.wikipedia.org/wiki/Nerf


Nerf in a game context was metaphorical, based on Nerf guns. Nerf guns' damage capability was a significant downgrade compared to a real gun. A previously powerful item in a game that was "Nerfed" was therefore a significant downgrade in capability.


That is what the preceding comment implies. After all, "nerf bats" directly references them. What was new, was simply the usage of making something worse.


It would have to for the comment to make sense.


Nerf the noun did, but not nerf the verb.


Although the nerfed Tesla batteries still charge faster than a brand new Bolt.


Sure, but a Model S costs around twice as much as a Bolt so you’d expect something for that money.


The Model S is much larger and faster than the bolt, so they definitely got some things for that money.


That's a good point. I think it only applies to the older Model S, and they may have paid 3-4x more than a Bolt, although the Bolt may not have been available when most of those were purchased. The 3 and newer S/X should be good.


Yep, they've already worked through any number of issues to achieve their level of performance. In fact they may have already worked through something similar to this particular issue: Those nerfed batteries in your link were nerfed by the same ~10% loss that GM has had to place on Bolts, implying it was the high charge causing both issues. Most telling, Teslas's issue was with batteries they stopped using 4 years ago. I don't know if that signifies a 4-year lead on GM, but they're certainly ahead.


They degraded in that 4 years(2-3 at time of incidents) time, and when put under stresses, they caught fire. Tesla vehicles are not configured to avoid stressing their packs, therefore users were not made aware of degradations.

Isn’t this simple enough and correct enough? I’m tired of their social media native ads. Let’s keep it simple.


Tesla doesn't have battery technology, yet. Panasonic does, as do other companies. And what is "his tech"?

It's going to be a while until EVs are mass market, and you seem to be making an implicit assumption that Tesla will be immune to competition in such a market.


They don't mean the actual cells but the tech supporting them. Layout of the cells, battery management system, battery cooling system, charging circuitry, motor controller, etc. These are the hard parts and they are custom Tesla designs.


But that's not battery technology, is it? It's what you said, everything else but the battery technology. Also, Tesla isn't the only company with that technology.


That is battery technology in my opinion. Technology involved in managing and and controlling battery systems.

Batteries these days are very reliant on their controllers to perform well and be reliable. You couldn't just jam a dumb battery bank into a Tesla and have it work right.


Well, you could do that, and then it would catch fire. No problemo.


Hah, yes. It's as a like to joke with my boss (who has a good sense of humor) when asked to do something faster than reasonable:

"Well, if you're willing to accept an incompetent result, there's any number of things I could do very badly. If you'd like, you can fire 10 people & I'd happily accept their 10 salaries if I'm allowed to be bad at their jobs. If you insist on competence though, it's going to take more time"


But that's not battery technology, is it?

If you wish to be the guy insisting "That's not batteries, that's battery packs!" nobody can stop you.

Not really a good look though.


You guys are terrible at hiding COI.


If I was a paid shill or investor for anyone selling batteries, battery packs, electric vehicles or anything related I'd have a lot more money in my bank account than I do. Probably wouldn't be driving a car that predates Tesla's IPO either.


If you're not willing to have a conversation on the merits without resorting to accusing the other person of deliberate lies, you shouldn't be on HN.

Right now you're basically saying that anyone who disagrees with your extremely narrow definition can't just be wrong, but also must be a self-interested shill for Tesla. Sure, there's probably TSLA holders or employees here that talk it up, but your inability to see anyone's comment that reflects well on Tesla as something other than self-interest instead of honest opinion is devolving your comments to the quality of conspiracy theories instead discussion.


Whats COI?


conflict of interest, they are implying they are TSLA holders or shorters


I’m implying employees. “Holders or shorters” is one of their template narratives


Just because those narratives may be used by the self-interested doesn't mean that are false in all cases. There are shorters, there are shills pushing the opposite way, but not every negative things said is a shorter, and not everything positive said is a shill.


You defended that narratives.


The chemistry and assembly of battery cells, and the system design of battery packs, are completely independent disciplines. I understand where the parent commenter is coming from - all the headlines make out that Tesla has some new battery (cell) tech. They don't - they just cleverly assemble Panasonic cells.


You cannot reduce "battery technology" simply to the cell level. Because the cell isn't really battery tech in that pedantic sense either. If you're going to say the things that make batteries functional in their applications aren't part of the tech, then you're left with defining batteries merely as the chemical reactions that take place. It's like saying the glass housing isn't part of light bulb technology because it's not the element that actually gives off light. That the legs on a chair are not pair of chair technology because it's not the part you actually sit on.

You cannot divorce the things that make something function in any practical sense for the concept of that area of technology.

You seem to be taking issue with this topic because some articles imply it's the cell that's been innovated. You have a point: The articles should be clearer. But the failure of those articles to properly make that distinction doesn't mean it's not part of battery technology.

Let's make this simpler: There is battery cell technology and there is battery management technology. Both are part of high tech implementations of battery systems.


You're absolutely correct, and it matters not one bit because for the general public (and even the not-so-general public of HN readers) they don't care that much about the difference and it really doesn't matter much if at all for the topic at hand.


Absurd. Charge management has everything to do with thermal management and service life optimization, which in turn has everything to do with packaging and chemistry.


They are not completely independent, the pack has to understand cell technology. That's like saying to develop a good cluster you have to understand nothing about individual servers.

And Tesla actually has new cell technology and new pack technology that has nothing to do with Panasonic.


That's kind of like saying software isn't part of the technology it controls. But the two are fairly inseparable.

If the software & control systems are specific & integral to the hardware they're attached to, it's part of the same package. If we were talking about a CNC laser cutter, the specialized software is also part of the package "CNC technology". You literally cannot have the first without the second. In Tesla's case, theirs seems to be better. Not that they haven't had problems: It's that they've already worked through a bunch of them, and as I understand it have come up with some novel innovations in the process.


That's like saying AMD doesn't have microprocessor technology, yet, TSMC does, as do other companies.

It's true that Tesla doesn't manufacture the batteries, but the technology behind the batteries is proprietary and exclusive to Tesla, atleast for the time being.


Tesla has battery tech. From what I've read, they've been licensing whatever bits are helpful (eg chemistry) to whoever produces their batteries, but they're preparing to ramp production of in house cells over the next few years.

https://fortune.com/2020/09/21/tesla-battery-day-what-to-exp...


Of course they do. Specifically they went into a lot of details during last investor presentation.


They announced plans. That's different than has or have.


Oh, come on. Yes, Panasonic carries part of the partnership when it comes to battery cells manufacturing. Tesla is still heavily involved in decisions regarding cell chemistry, and exclusively involved regarding battery pack design and manufacturing. Avoiding fires, charging quickl, ensuring battery longevity and making sure you get the most miles per kWh into the pack are largely battery pack issues.

But then again, the longer this confusion lasts, the longer before it becomes obvious to everyone that Tesla has a significant in-house lead on this technology, so meh.


If you want do be nitpicky about it, they announced specific cell designs, which means that they developed them, i.e. they have the “technology“.


If you think the cells are the extent of the battery tech, why don’t you just consider TFA?


Yes, I don't think GM would get very far saying that their batteries causing fires has nothing to do with battery technology.


Sounds like a software update. Maybe putting wifi and remote updating in the first place was the cheaper/better idea after all.


GM cars don't need wifi. GM pioneered the connected car via cellular modems through their OnStar program since 1996...

Question is how much of your vehicle software do you expose.


> Dealerships will update the vehicle’s battery software beginning next week.

Sound like you get to drive it in to the dealership anyway.

If I remember correctly some of the dealership agreements mandate that the dealer handles software updates preventing some manufacturers from legally being able to OTA without dealer involvement.


Another point seems to be upgrade fees - e.g. with Volvo I have to pay for the dealers diagnostics software subscription, for the actual software update itself (yes, really), and for the time it takes because it blocks a slot in the dealers garage (and the diagnostics system). Upgrades effectively easily cost 700 Euro per year, and they can't even tell me what's actually changed (at least it's optional).

If we only had the technology to roll out tamperproof updates, even if it was via a USB stick... /s (At least the map updates happen exactly that way).

Ergo, I'm under the impression they're really just only using this to siphon more money from the customers.


This is for non-mandatory updates - things like navigation and updates to a slightly less crappy UI.

For safety recalls the dealer bills the manufacturer for their time, so as the customer you don't need to pay anything (other than your time to take the car to the dealer).

I have a 3rd hand Toyota, and after I bought it I noticed none of the recalls had been done. I called up my local dealer, they ordered the parts, then two weeks later they replaced airbags, fuel tank lid and updated the ECU without me having to pay anything.


Oh, the paid updates might affect the infotainment system, but might also affect the ECU instead - at least as far as I'm aware, and then of course this might vary depending on the model year. Updates as part of a recall are free.


FCA will let you do your own software updates as long as you buy their micropod. I'll have to look up which ones require dealerships because thats a bit strange to me. A lot of dealerships have a guy they call to come do flashes for them. Usually an independent contractor that circles around dealerships of every make flashing things for everybody.


That sucks...

but at the other end of the spectrum, Tesla have had issues with cars running into walls while the driver was enjoying self driving OTA content, or cars catching fire after new fast charging OTA content, in neither of the cases they issued recalls as legally defined.

So I’m not so sure whether manufacturers should be allowed direct OTA updates without recalls or heavy regulatory oversights.

Boeing could have designed MCAS on 737 MAX so that the company can rewrite its behaviors covertly and remotely, except that that isn’t how airplanes work. Why should it be how cars work?


> I’m not so sure whether manufacturers should be allowed direct OTA updates without recalls or heavy regulatory oversights.

I think there probably should be some oversight, but not all ECU updates are serious enough to warrant an official recall. Sometimes they're just cosmetic changes or improvements to the head unit's user interface. Sometimes they're bug fixes for convenience features that are not a risk to the operation of the vehicle. Sometimes they're compliance related (ex. Volvo updated my headunit so that the 360 surround camera view couldn't be the default view when shifting into reverse (a US law requires that the backup camera is seen when shifting into reverse).


Onstar was created to make money and its damn expensive. they really push it hard when buying any GM product. My Volt advisor at the time did a good job of explaining how it benefited him and the dealership and how long I would need to keep it active to help them; which was not beyond the free period so it was no big deal to me.

what got annoying was more than once Onstar tried to put some app functionality behind a paywall. This included even the ability to check air pressure (last reading) and then remote start.


This is just a temporary fix, limits the maximum charge level. They're still investigating the issue and coming up with a proper fix, they just know that it's happening to fully charged cars.

For what it's worth, I popped out to our Bolt this afternoon after getting the email, and tweaked the max charge level to 90% from the panel. Easiy done.


Wait, shouldn't you be charging to 80% in normal circumstances anyway, because keeping the battery at 80%+ for long periods of time deteriorates it faster?


We use the Bolt for almost all our. It doesn't stay above 90% for any great length of time.


Yes. I think the real story here has nothing to do with fire risk, but that GM needs to drag 68,677 cars into the shop because they're too incompetent to master secure over the air updating of software. If a "risk" doesn't meet the metric of needing a recall I guess GM is just not going to fix it. If a brand already has over-the-air updating in place, inadequacies tend to get fixed big and small.


they should also update the documentation:

https://en.wikipedia.org/wiki/Halt_and_Catch_Fire_(computing...


Alternatively, a technically valid but evil solution is updating only the documentation. Much cheaper to implement. Any hardware engineer will understand what does it mean by that ;-)

A feature is broken? Just update the documentation and say it's broken, problem fixed! No need to fix the actual problem... An even more evil variant: still claim the feature works in the datasheet, but hide the problems in the erratas.


You cannot fix a chemistry error over the Internet


You can if the chemistry mistake causes problems if you charge too fast or charge to too high a voltage. Both can be avoided (fixed?) if the OTA update allows you to control the charging.


I would like to know more details. Specifically, what changed in 2020 that makes it no longer a risk? I ask because I have a 2020 Bolt sitting out in my driveway.


Only batteries made by LG Chem have had the issue. I guess they changed suppliers.


They didn't change suppliers they changed the chemistry to provide more range [1]

[1] https://www.sae.org/news/2019/12/chevy-tweaks-2020-bolt%E2%8...


LG Chem provide the batteries and the entire powertrain, as well as the infotainment system. Chevy is pretty committed to them for this car.


My Bolt has been at the dealership for last 2 weeks after the battery failed. It failed while backing out of driveway after the first full charge in a month or two. I was planning to pick it up tomorrow, but now I may delay that.


10% instant range reduction for Bolt EVs.


For now, but likely everyone will end up with a nice new battery. Any less than original range and I can see some massive lawsuits coming.


It sounds like some law firm is going to make $100 million and the vehicle owners are going to get a $100 certificate to use at their next service appointment.


Unfortunately you’re probably correct on that. Class action suits don’t really seem to provide any benefit to the end user.


Hyundai just did something similar with their Kona EVs last month.


At least they don't explode on rear impact at medium/low speeds: https://www.tortmuseum.org/ford-pinto/#:~:text=Grimshaw%20v.....


I wonder how Cruise Automation is affected by having their self-driving tech running on the Bolt platform.


Probably not much:

> GM said it has developed software that will limit vehicle charging to 90% of full capacity to mitigate the risk while GM works to determine the appropriate final repair.


I'm not sure if Bolts have Over The Air updates? It mentions dealers...I know some GM cars have or will have...just not sure the older Bolts.

So a consumer would need to personally ensure they don't charge to 90%, or bring it back to the dealership twice.


The bolt has over the air update capabilities, but iirc, it was used only once in the past 4 years. And the bolt has an option to limit charge to X%. I set mine to about 80%, and it's plenty for my needs. So you can leave your car plugged in all day long


Huh. I only charge my Tesla to 100% a few times a year.


I would guess they don't charge over 90% anyway.

I've only charged our 3 to 100% 5 or 10 times over 2.5 years. And I don't leave it sitting that way.


I've been shopping for a Bolt on and off for a couple months and you'd be surprised! I see folks all the time that say some dismissive equivalent of "oh it's a lease, I just charge to 100% every day". Not charging to full capacity will either become car knowledge that everyone learns in drivers ed or manufacturers will end up blocking off some percentage in software to prevent full charges.


Tesla has a slider for the charge limit with 90%+ being designated as Trip and 50-90% as Daily. Even for a trip, I avoid charging above 90% because regenerative braking stops working when the battery is nearly full, which results in a substantial change to the way the car handles. You get used to one pedal driving and rarely having to use the brakes, so it's very unnerving for the car to not slow down rapidly when you let up on the accelerator pedal.


Frankly I would prefer the software cap which I can disable (with a warning) for long trips. I like to plug in overnight and I can't check at 2 AM where the charge is at.


That's (basically) what the charge limit already does; nobody's running out to their car at 2am to unplug it before it hits 100%.

Presumably these "it's a lease" folks are just choosing to set the 'limit' at 100%.

I agree with other comments that I prefer the regen anyway; it's weird to charge to 100% and have to use the brakes to slow the car so much. Your efficiency on that last 10% of charge is lower because of this lack of regen.


The charge limit is not configurable on my car. Not sure about the Bolt or Teslas, though.


It's configurable on a Tesla.


First, every EV battery already has excess capacity you can't charge up.

Second, Nissan tried having the Leaf default to 85%. But they got in trouble for claiming more range than they should. The government rates the capacity in the default configuration.


There’s no reason you couldn’t have the default be 90% of whatever you thought was safe and allow 11% over rated on occasion by explicit user selection. It’d be like packing an extra gallon or two into your gas tank after the dispenser clicks off.


This was annoying because now my Leaf charges to 100% all the time, even when I never need to do that…


Yeah it makes no sense for the operator to manually regulate charging. We don't do that with any other rechargeable device. Oh gotta unplug my phone it's at 90%!


Exactly, the gas pump stops when the tank is full, not when its 10% overfilled.


> I see folks all the time that say some dismissive equivalent of "oh it's a lease, I just charge to 100% every day"

Guilty as charged :). We charge our Model 3 to 80%. But my wife's Bolt is a lease, so we just left it at the default so it always has maximum range.


Wait, if this is a known safety issue with batteries, why isn't there an engineered capacity head room? Set the max charge to the cells' 90% capacity, but report 100%.


That's what they already do. With lithium batteries neither 100% nor 0% are at the physical limits of the cell, they are both software limits. They might be enforced at the battery pack (over/undervoltage protection) or higher up in the chain.

It sounds like GM set the level too high, but it could also be something else like a problem in the battery's thermal management that is fixed by limiting the maximum charge level.


It's a known long-term-battery-health issue. The catching-on-fire part is not normal.


The Model 3 has a setting to do that. The Bolt does not. If you plug it in, it's charging to full.

With the Bolt, all you can do is use tricks like scheduling the charging late (so that it might not be fully charged by the time you drive off in the morning).


Given GM's history with the EV1, I'm not sure if I'd be returning my car to the dealer to have this recall applied: "software that will limit vehicle charging to 90% of full capacity to mitigate the risk while GM works to determine the appropriate final repair."


GM is leaning heavily into electric. This is not the 90s. They are not going to scrap your car.


If it requires cellular updates to prevent starting on fire is the term your car applicable?

I would never buy one of these new botnet cars.


Well…wait, how is that different than having a non-botnet car with a serious defect?

Say, all those cars with defective Takata airbags. I guess you’re arguing that you feel safer that the manufacturer can’t remotely disable the car, but I don’t see how how a cell connection has much to do with a car having a deadly issue or not. It has to get fixed no matter what.

(Also, given how many cars are leased, and the increasingly long time of those leases…lots of people, in fact, do not own their cars.)


Non botnet cars have a limited amount of slack in the rope left. Fact is most cars now require propriety expensive systems just to "repair" More so it is about how software in cars is pushed out to remedy bad design and engineering in the first place.

Idealistic I suppose, but what power exist on the consumers side? How much of a reach is to say modern design and engineering has turned into a cat a mouse game. One in which we should seriously consider how deeply cancerous the lack of ownership on vehicles is. Before accepting it for common place we should look at the way our incentives have been shaped by the manufacturer.

Now, a vehicle could start on fire somewhere parked without access to the update for N reasons and they just need to prove the update was prevented by the owner and the result is also therefore attributable.

We can't get cell phone companies to update alot of phones past a year or so. How is GM going todo?

Constantly


Given 20 years difference and the known reasons why the EV1 leases were ended on time, it doesn’t really seem relevant to compare to Bolts that have been sold to customers.


The known reason being GM didn't want to invest in next generation technology? Or what are the 'known reasons'?


https://en.wikipedia.org/wiki/General_Motors_EV1#Cancellatio...

> In late 2003, General Motors, then led by CEO Rick Wagoner, officially canceled the EV1 program.[11][46] GM stated that it could not sell enough of the cars to make the EV1 profitable.[47] In addition, the cost of maintaining a parts supply and service infrastructure for the 15-year minimum required by the state of California meant that existing leases would not be renewed, and all the cars would have to be returned to GM's possession.

Obviously, market conditions 20 years later are very, very different and with increasing government mandates for zero emissions car sales, maintaining and improving on sales of their EVs is a key component of GM’s future in a way that it couldn’t have been in 2001.

Edit: also, to point out this article is about 69,000 Bolt EVs - only about 1000 EV1s were ever produced. The difference in scale and investment is massive.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: